The effects of Impact Factor


IN last month's editorial I looked at why people publish (Dunn 2006), here I consider one of the determinants of where articles appear.

Ideally authors should target the material they produce to an appropriate audience by publishing in different journals according to the subject of their research. Scientific journals want to publish high quality research that has been carefully peer-reviewed and is relevant to their subscribers. The result should be a mutually satisfying relationship for the editor, author and the reader. Sadly the balance is often disturbed by an artificial concept - the Impact Factor.

Why do journals care about impact factor?

The past impact factor of a journal has a knock-on effect on the quality of the submissions it receives. Authors from academic institutions can become obsessed with impact factor which is increasingly used to judge the quality of the output of researchers and their groups. Assessment of academic staff for promotion, pay rises, budgets, or grant proposals is often based on the impact factor of the journals in which they publish. To improve their personal rating, academics naturally try to publish in the journals with the highest impact factors. This means that papers from academic institutions (which should describe groundbreaking research and may therefore be more frequently cited) are published in journals with high impact factors.

In times of cutbacks on academic funding libraries have to make difficult decisions about which publications they hold. In some cases this decision is made (at least in part) on the basis of impact factor.

What is impact factor?

The Impact Factor was first proposed by Garfield in 1955 and introduced as a measure in the early 1960s. Today journals are often judged by their Impact Factor - a measure of the frequency with which the ‘average article’ in the journal has been cited in a particular year or period. It is calculated by summing the number of citations to articles published in the journal over the past two years and dividing by the total number of papers published in this time.

Articles that are infrequently cited ‘dilute’; the effects of other more popular articles and reduce the impact factor of the journal. Most journals try to increase their impact factor by publishing more articles that are likely to be cited and reducing the number of those that may receive few citations (usually case reports).

Although, the impact factor gives an indication of the importance of articles published in a given journal it provides no information about the merits of individual papers and tells us little about clinical relevance. Impact factors correlate poorly with the number of citations received by individual articles. The most cited 50% of articles in a publication are cited ten times more often than the rest (Seglen, 1997). Perhaps a better measure for individual papers would be the number of times they are actually cited.

Many believe it would be better if authors were able to accumulate a personal Citation Index - reflecting the number of citations received by all their papers. Although technology (with online publication) permits this form of scoring the concept has not been well received by many authors – perhaps it is just too personal…

What effect does impact factor have on readers?

Impact factor has historically also been indirectly linked with subscriber numbers – papers in a journal with a large readership were more likely to be cited than smaller (particularly foreign language) journals. The increasing presence of journals on searchable online databases such as PubMed has reduced this effect. However, the increasing reliance on online publication puts more pressure on journals to have a widely accessible website. It is still true that the majority of the veterinary journals with the highest impact factors are American and much of the high quality research performed in the UK is published in these journals and may be lost to the UK audience.

Many journals do not accept submission of reports of single cases. The reasons for this are diverse but certainly many editors are influenced by the negative effect case reports can have on a journal's impact factor. By their nature case reports describe rare or unusual cases and it is therefore unlikely that there will be many future papers on the same subject that will refer back to the original work. This lack of citations affects the impact factor of the journal as a whole.

The future

A high impact factor is always hard to achieve for a clinical journal. The articles in these journals may not be frequently cited in the short term but clinical journal editors tend to care more about whether the articles are read and for how long they continue to have clinical relevance. Analysis of traffic on the new JSAP website should allow us to monitor the frequency with which articles are downloaded or accessed and to identify the most popular papers amongst readers.

Papers in high impact factor journals do not achieve more citations than if they had been published in a lower ranking journal (the so-called free-ride hypothesis). So choosing to submit a paper to the journal with the proper audience should be one of the major concerns for authors. If a personal citation index were adopted authors would be completely free of the effects of impact factor – their publications would stand or fall by their own merits. Authors would then be able to select journals based on factors such as time to publication, quality of peer-review process and appropriate audience.

Hopefully the wheel is starting to turn in the favour of UK practitioner-focused journals. An increasing number of high quality submissions are being received from non-academic institutions especially the private referral centres. The authors of these submissions have no external pressures to publish in journals with high impact factors. For these authors selecting an appropriate audience for their work becomes the most important criterion in choosing the journal in which to publish. Publishing in the JSAP targets a large proportion of UK small animal vets (the BSAVA membership). Until the number of high quality submissions from general practice increases significantly it is essential for the JSAP to strive for as high an impact factor as possible – without this we cannot guarantee access to the best submissions from academia.